Research

Very smart students whom I am fortunate to supervise in their thesis

- Can Pouliquen, Nov 2022 - present -

Co-supervision with Mathurin Massias and Paulo Gonçalves (HDR).
Automatic estimation of functional connectivity graphs in neuroscience.
Objective: Develop machine learning methods for faithfully, rapidly, and without calibration estimating dynamic graphs from M/EEG signals.

- Hugues Van Assel, Sept 2021 - present -

Co-supervision with Aurélien Garivier (HDR).
Latent graphical models for dimensionality reduction.
Objective: Develop new dimensionality reduction methods using tools from optimal transport, graph analysis, and probabilistic modeling.

Some workshops

• We co-organized with Mathurin Massias of the RT MIA thematic day entitled ‘‘Dimensionality Reduction for Learning and Visualization’’ (November 10, 2023). Link to the program.
• In partnership with IXXI, I organized the day entitled ‘‘Frugality and Machine Learning’’ (September 11, 2023). Link to the program.
• We co-organized with Arnaud Breloy the special session entitled ‘‘Graph Learning and Learning with Graphs’’ during GRETSI 2023.

Some selected talks

• Unifying Dimensionality Reduction and Clustering with Gromov-Wasserstein Projection.
• CIRM November 2022 Towards Compressive Recovery of Sparse Precision Matrices.
• European Summer School on AI (IDESSAI 2022): From compressed sensing to machine learning to compressive learning.
• November & December 2021: I gave a talk “Less is More ? How Optimal Transport can help for compressive learning” in CMAP Ecole Polytechnique for the GDR ISIS meeting Transport Optimal et Apprentissage Statistique
  . You can find the slides here.
• January 2020: I was invited at MIA UMR 518 Paris to talk about “The optimal transportation problem for structured data”. You can find the slides here.